Genetic algorithm optimization and realization of broadband wire antennas.

摘要

In this dissertation is presented a compilation of papers which focus upon the development and realization of broadband monopole-type and dipole-type antennas. Although monopoles and dipoles in their simplest forms are inherently narrowband, they have radiation characteristics, such as linear polarization and omnidirectional azimuth-plane patterns, which are desirable for ground-based mobile communication systems. Genetic algorithms (GA) and integral equation solution techniques with a lumped-load model included are used to determine the location and component values of load circuits and the values of matching network elements leading to bandwidth improvement in straight-wire antennas. Although most of the effort of this dissertation has been devoted to the development of theoretical and numerical tools for designing broadband antennas, laboratory models have been fabricated and their performances have been evaluated experimentally. Bandwidths of 5:1, 12.6:1, and 20:1 for voltage standing wave ratio (VSWR) less than 3.5 and system gain greater than −4.5 dBi are achieved for constructed monopole antennas loaded with parallel inductor-resistor circuits and having transmission-line-transformer matching networks. Since the loads and matching networks contain losses, the trade-off in bandwidth and efficiency is key to developing these systems. Genetic algorithms are also applied to increase bandwidth of sleeve-cage and sleeve-helical antennas via optimization of the wire parasites. The absence of loads and matching networks in these systems eliminates losses at the expense of greater bandwidth potential. Measurements of VSWR and input impedance confirm the bandwidth improvement of these antennas.